A mechanistic investigation of the photoinduced reduction of carbon

Jose F. Martinez , Nathan T. La Porte , and Michael R. Wasielewski ... Luke A. Adams , Nicholas Barlow , Stefano Ferrari , Bim Graham , and Gilles Gas...
1 downloads 0 Views 784KB Size
Organometallics 1985, 4, 2161-2166

2161

A Mechanistic Investigation of the Photoinduced Reduction of Carbon Dioxide Mediated by Tricarbonylbromo(2,2’-bipyridine)rhenium( I ) Charles Kutal,*+ Michael A. Weber,+ Guillermo Ferraudi,* $ and David Geiger$ Department of Chemistty, University of Georgk Athens, G8orgia 30602, and the Radiation Laboratory, University of Notre Dame, Notre Dame, Indiana 46556 Received February 15, 1985

A mechanistic study of the photoinduced reduction of carbon dioxide to carbon monoxide in the triethanolamine/dimethylformamide/ReBr(C0)3(bpy) (bpy is 2,2’-bipyridine) system is described. Continuous photolysis at 436 nm results in the highly specific formation of CO with a quantum yield, +cot that reaches 0.15. The value of +co decreases with increasing triethanolamine concentration in the range 0.75-3.8 M; addition of 10% water to the solvent medium also lowers q 5 c ~ . Luminescence measurements reveal that triethanolamine reductively quenches the Re-to-bpy charge-transfer excited state of ReBr(CO),(bpy) with a rate constant of 6 X lo7 M-’ s-l, whereas COz undergoes no discernible interaction with the photoexcited complex. Formation and decay of the initial reduction product [ReBr(CO),(bpy)]- have been observed in flash photolysis experiments. The amount of [ReBr(CO),(bpy)]- produced correlates with the value of q5co, thereby implicating this 19e complex in the mechanism of C02 reduction. Evidence that [ReBr(CO)3(bpy)]- reacts directly with COz has been obtained, although the identity of the resulting product is unknown at present. Introduction Nonbiological systems capable of reducing carbon dioxide are of considerable interest as a means of transforming this abundant raw material into fuels and organic Recently, Lehn and co-workers reported3c that visible-light irradiation of ReA(CO)3(bpy) (A is C1 or Br; bpy is 2,2’-bipyridine) in a C02-saturated solvent mixture of triethanolamine (TEOA)/dimethylformamide (DMF) results in the reduction of carbon dioxide to carbon monoxide with high chemical specificity. It was proposed that reductive quenching of the lowest Re-to-bpy chargetransfer excited state of the complex by TEOA generates a species, [ReA(CO),(bpy)]-, capable of mediating the two-electron process represented by eq l.4 While few

COz + 2H++ 2e- = CO + HzO

(1)

details concerning the nature of the interaction between this reduced species and COz are currently available, it seems clear that a great deal of interesting and potentially useful chemistry is occurring at the Re center. We have undertaken a mechanistic investigation of the ReBr(CO),(bpy)/TEOA/DMF/CO, system and report here luminescence quenching, continuous photolysis, and flash photolysis data. Our results confirm the proposal concerning the initial photochemical act and also provide additional insight about the overall mechanism of COz reduction.

Experimental Section (a) Reagents. Dirhenium decacarbonyl (Aldrich)was purified by vacuum sublimation at 95 O C . Addition of a slight excess of bromine to a chilled (5 “C) methylene chloride solution of Rez(CO),, yielded ReBr(CO),; evaporation of the solvent afforded the solid compound which was washed with distilled water and air-dried. Reaction of ReBr(CO)5and bpy according to the directions of Wrighton and Morse5gave ReBr(CO)3(bpy)(reported to be the facial isomer) in good yield. The purity of the product was established by spectroscopy (UV-visible and infrared) and elemental analysis. ‘University of Georgia. t

Notre Dame Radiation Laboratory.

0276-7333/85/2304-2161$01.50/0

Dimethylformamide (Aldrich Gold Label) and triethylamine (Baker reagent grade) were dried over potassium hydroxide and then diskilled from calcium oxide. Triethanolamine (Aldrich)was used as received. The carbon dioxide (&lox) employed in the photochemical studies was found to be 299% pure by gas chromatography. (b) Physical Measurements. Electronic absorption spectra were recorded on Cary 15 and 219 spectrophotometers. Infrared spectra were measured on a Perkin-Elmer 599B spectrometer. Luminescence spectra were taken with a Perkin-Elmer MPF-44B spectrofluorimeter and are uncorrected for photomultiplier response. Lifetimes were determined with an Ortec 9200 timeresolved spectrometer. Viscosities were measured with a Cannon viscometer. (c) Continuous Photolysis Procedures. Photolyses at 436 nm were conducted with a 200-W high-pressure mercury-arclamp (Illumination Industries) in conjunction with suitable interference and blocking filters. Incident light intensity was determined by Reineckate actinometry! In a typical photochemical run, 3 mL of a TEOA/DMF solution containing ReBr(CO),(bpy) was placed into a 1-cm rectangular quartz cell fitted with a rubber septum. The solution was bubbled for 15-30 min with either Ar or COz, placed in a thermostated cell holder, and then irrediated while being stirred. A known volume of gas from the headspace above (1)For reviews of transition metal-C02 chemistry, see: (a) Eisenberg, R.; Hendriksen, D. E. Ado. Catal. 1979,28,79.(b) Sneeden, R.P. A. in “Comprehensive Organometallic Chemistry”; Wilkinson, G., Ed.; Pergamon Press: Oxford, 1982;Vol. 8,Chapter 50.4. (c) Ito, T.;Yamamoto, A. In “Organic and Bio-organic Chemistry of Carbon Dioxide” Inoue, S., Yamazaki, N., Eds.; Wiley: New York, 1982;Chapter 3. (d) Darensbourg, D. J.; Kudaroski, R. A. Adu. Organomet. Chem. 1983,22,129. (2) For recent studies of electrochemical CO, reduction, see: (a) Fischer, B.; Eisenberg, R. J. Am. Chem. Sac. 1980,102,7361. (b) Amatore, C.; Saveant, J.-M. Zbid. 1981,103,5021. (c) Tezuka, M.; Yajima, T.; Tsuchiya, A.; Mataumoto, Y.;Uchida, Y.; Hidai, M. Zbid. 1982,104,6834. (d) KaDusta. S.:Hackerman. N. J. Electrochem. Sac. 1983.130.607. (e) Liebe< C. M.;.Lewis, N. S: J. Am. Chem. Sac. 1984, 1b6,5033. (fj Hawecker. J.: Lehn. J.-M.: Zieasel. R. J . Chem. SOC..Chem. Commun. 1984,328.‘ ’ (3)For recent studies of photochemicaI or photoelectrochemical CO, reduction, see: (a) Lehn, J.-M.; Ziessel, R. Proc. Natl. Acad. Sci. U.S.A. 1982,79,701. (b) Bradley, M.G.; Tysak, T.; Graves, D. J.; Vlachopoulos, N. A. J. Chem. Soc., Chem. Commun. 1983,349. (c) Hawecker, J.; Lehn, J.-M.;Ziessel, R. Zbid. 1983,536.(d) Halmann, M.; Katzir, V.; Borgarello, E.; Kiwi, J. Sol. Energy Mater. 1984,10, 85. (e) Hawecker, J.; Lehn, J.-M.; Ziessel, R. J. Chem. SOC., Chem. Commun. 1986,56. (4) The redox potential for the process given by eq 1 is -0.52 V vs. NHE in a pH 7 aqueous solution. (5)Wrighton, M.; Morse, D. L. J. Am. Chem. Sac. 1974, 96,998. (6)Wegner, E.;Adamson, A. W. J. Am. Chem. Sac. 1966,88,394.

0 1985 American Chemical Society

2162 Organometallics, Vol. 4 , No. 12, 1985

Kutal et al.

4

3

10/T 2

I

OS2

0.4

0.6

0.0

1.0

[TEOA], M Figure 1. Stern-Volmer plot for quenchin of the emissive excited state in ReBr(CO),(bpy) by TEOA; $and I denote luminescence intensity in the absence and presence of quencher, solution bubbled respectively: ( 0 )solution bubbed with Ar; (0) with COz. the photolyzed solution was removed with a syringe (inserted into the cell through the rubber septum) and analyzed for CO and H2 by gas chromatography. The equipment and procedures employed in this analysis are described el~ewhere.~ A systematic error occurs in the measurement of the quantum yield for GO production, since no account is taken of the CO that remains dissolved in the sample solution. The magnitude of this error can be estimated as follows. The solubility of CO in DMF at 25 "C and 1 atm gas pressure is 2.5 x M,8 and from the ideal gas law, the concentration in the gas phase under these conditions is 4.1 X M. Thus the equilibrium constant for partitioning of CO between the two phases (eq 2) is 16.4. Given CO(DMF)

CO(gas)

(2)

that the sample solution and gaseous headspace in our photolysis cells are of comparable volume, a straightfornard calculation shows that >94% of the CO produced during irradiation should reside in the gas phase. Consequently, the error in the quantum yield attributable to dissolved CO is minor (